Harald Hoyer
42c52bd87f
The bot no longer shells out to `opencode run`. Instead it POSTs to the OpenAI-compatible /chat/completions endpoint exposed by llama-server on halo.hoyer.tail:8000 directly. This removes the Bun/sqlite cold-start overhead per request, drops the pkgs.opencode runtime dependency, and eliminates the ExecStartPre dance that materialized config.json into the service's $HOME. Conversation history is now stored as a proper OpenAI `messages` list with system/user/assistant roles, instead of the XML blob that was inlined into a single `opencode run` argument. The interactive opencode setup (config/opencode/config.json) is unchanged — only the bot stops depending on it. The module gains a `modelBaseUrl` option; `model` is now the bare model name (`halo-8000`) without the provider/ prefix that the opencode CLI required.
320 lines
10 KiB
Python
320 lines
10 KiB
Python
#!/usr/bin/env python3
|
||
"""
|
||
Nextcloud Talk OpenCode Bot
|
||
|
||
Receives webhooks from Nextcloud Talk and forwards the conversation to an
|
||
OpenAI-compatible chat-completions endpoint (e.g. llama-server) running on
|
||
the local LLM host.
|
||
"""
|
||
|
||
import hashlib
|
||
import hmac
|
||
import json
|
||
import logging
|
||
import os
|
||
import re
|
||
import secrets
|
||
from datetime import datetime
|
||
from typing import Optional
|
||
|
||
import httpx
|
||
from fastapi import FastAPI, Request, HTTPException, Header
|
||
from fastapi.responses import JSONResponse
|
||
|
||
NEXTCLOUD_URL = os.environ.get("NEXTCLOUD_URL", "").rstrip("/")
|
||
MODEL_BASE_URL = os.environ.get("MODEL_BASE_URL", "").rstrip("/")
|
||
MODEL_NAME = os.environ.get("MODEL_NAME", "halo-8000")
|
||
MODEL_API_KEY = os.environ.get("MODEL_API_KEY", "")
|
||
ALLOWED_USERS = [u.strip() for u in os.environ.get("ALLOWED_USERS", "").split(",") if u.strip()]
|
||
TIMEOUT = int(os.environ.get("TIMEOUT", "120"))
|
||
SYSTEM_PROMPT = os.environ.get("SYSTEM_PROMPT", "")
|
||
BOT_NAME = os.environ.get("BOT_NAME", "Halo")
|
||
|
||
|
||
def get_bot_secret() -> str:
|
||
cred_path = os.environ.get("CREDENTIALS_DIRECTORY", "")
|
||
if cred_path:
|
||
secret_file = os.path.join(cred_path, "bot-secret")
|
||
if os.path.exists(secret_file):
|
||
with open(secret_file) as f:
|
||
return f.read().strip()
|
||
return os.environ.get("BOT_SECRET", "")
|
||
|
||
|
||
BOT_SECRET = get_bot_secret()
|
||
|
||
logging.basicConfig(
|
||
level=logging.INFO,
|
||
format="%(asctime)s [%(levelname)s] %(message)s"
|
||
)
|
||
log = logging.getLogger(__name__)
|
||
|
||
app = FastAPI(title="Nextcloud OpenCode Bot")
|
||
|
||
# Conversation history per room: list of OpenAI-style message dicts
|
||
# (role: "user"|"assistant", content: str).
|
||
conversations: dict[str, list[dict]] = {}
|
||
MAX_HISTORY = int(os.environ.get("CONTEXT_MESSAGES", "6"))
|
||
|
||
|
||
def generate_bot_auth_headers(body: str = "") -> dict:
|
||
random = secrets.token_hex(32)
|
||
digest = hmac.new(
|
||
BOT_SECRET.encode(),
|
||
(random + body).encode(),
|
||
hashlib.sha256
|
||
).hexdigest()
|
||
return {
|
||
"X-Nextcloud-Talk-Bot-Random": random,
|
||
"X-Nextcloud-Talk-Bot-Signature": digest,
|
||
"OCS-APIRequest": "true",
|
||
}
|
||
|
||
|
||
def verify_signature(body: bytes, signature: str, random: Optional[str] = None) -> bool:
|
||
if not BOT_SECRET:
|
||
log.warning("No bot secret configured, skipping signature verification")
|
||
return True
|
||
|
||
if signature.startswith("sha256="):
|
||
signature = signature[7:]
|
||
|
||
expected1 = hmac.new(BOT_SECRET.encode(), body, hashlib.sha256).hexdigest()
|
||
if random:
|
||
expected2 = hmac.new(BOT_SECRET.encode(), (random.encode() + body), hashlib.sha256).hexdigest()
|
||
else:
|
||
expected2 = None
|
||
|
||
if hmac.compare_digest(expected1, signature):
|
||
return True
|
||
if expected2 and hmac.compare_digest(expected2, signature):
|
||
return True
|
||
return False
|
||
|
||
|
||
BOT_SYSTEM_PROMPT = """\
|
||
Du bist ein KI-Assistent im Nextcloud Talk Chat.
|
||
Deine Antworten werden direkt in den Chatraum gepostet.
|
||
Halte deine Antworten kurz und prägnant, da es ein Chat ist.
|
||
Nutze Markdown für Formatierung wenn sinnvoll."""
|
||
|
||
|
||
def build_system_prompt() -> str:
|
||
if SYSTEM_PROMPT:
|
||
return f"{BOT_SYSTEM_PROMPT}\n\n{SYSTEM_PROMPT.strip()}"
|
||
return BOT_SYSTEM_PROMPT
|
||
|
||
|
||
def build_messages(conversation_token: str, current_message: str, current_user: str) -> list[dict]:
|
||
messages: list[dict] = [{"role": "system", "content": build_system_prompt()}]
|
||
history = conversations.get(conversation_token, [])
|
||
messages.extend(history[-MAX_HISTORY * 2:])
|
||
messages.append({"role": "user", "content": f"[{current_user}] {current_message}"})
|
||
return messages
|
||
|
||
|
||
async def call_model(messages: list[dict]) -> str:
|
||
"""POST to /chat/completions and return the assistant content."""
|
||
if not MODEL_BASE_URL:
|
||
return "❌ Fehler: MODEL_BASE_URL ist nicht konfiguriert."
|
||
|
||
url = f"{MODEL_BASE_URL}/chat/completions"
|
||
headers = {"Content-Type": "application/json"}
|
||
if MODEL_API_KEY:
|
||
headers["Authorization"] = f"Bearer {MODEL_API_KEY}"
|
||
|
||
payload = {
|
||
"model": MODEL_NAME,
|
||
"messages": messages,
|
||
"stream": False,
|
||
}
|
||
|
||
log.info(f"Calling model {MODEL_NAME} at {url} ({len(messages)} messages)")
|
||
|
||
try:
|
||
async with httpx.AsyncClient(timeout=TIMEOUT) as client:
|
||
resp = await client.post(url, json=payload, headers=headers)
|
||
|
||
if resp.status_code != 200:
|
||
log.error(f"Model API error: {resp.status_code} {resp.text[:500]}")
|
||
return f"❌ Fehler vom Modell: HTTP {resp.status_code}"
|
||
|
||
data = resp.json()
|
||
choices = data.get("choices") or []
|
||
if not choices:
|
||
log.error(f"Model returned no choices: {data}")
|
||
return "❌ Fehler: Modell hat keine Antwort geliefert."
|
||
|
||
content = choices[0].get("message", {}).get("content", "")
|
||
if not content:
|
||
log.error(f"Model returned empty content: {choices[0]}")
|
||
return "❌ Fehler: leere Antwort vom Modell."
|
||
return content.strip()
|
||
|
||
except httpx.TimeoutException:
|
||
log.error(f"Model API timeout after {TIMEOUT}s")
|
||
return f"⏱️ Timeout: Das Modell hat nicht innerhalb von {TIMEOUT}s geantwortet."
|
||
except Exception as e:
|
||
log.exception("Error calling model")
|
||
return f"❌ Fehler: {str(e)}"
|
||
|
||
|
||
async def send_reply(conversation_token: str, message: str, reply_to: int = None):
|
||
if not NEXTCLOUD_URL:
|
||
log.error("NEXTCLOUD_URL not configured")
|
||
return
|
||
|
||
url = f"{NEXTCLOUD_URL}/ocs/v2.php/apps/spreed/api/v1/bot/{conversation_token}/message"
|
||
|
||
headers = generate_bot_auth_headers(message)
|
||
headers["Content-Type"] = "application/json"
|
||
|
||
payload = {
|
||
"message": message,
|
||
"referenceId": hashlib.sha256(
|
||
f"{conversation_token}-{datetime.now().isoformat()}".encode()
|
||
).hexdigest()[:32],
|
||
}
|
||
|
||
if reply_to:
|
||
payload["replyTo"] = reply_to
|
||
|
||
async with httpx.AsyncClient() as client:
|
||
try:
|
||
resp = await client.post(url, json=payload, headers=headers)
|
||
if resp.status_code not in (200, 201):
|
||
log.error(f"Failed to send reply: {resp.status_code} {resp.text}")
|
||
else:
|
||
log.info(f"Reply sent to conversation {conversation_token}")
|
||
except Exception:
|
||
log.exception("Error sending reply to Nextcloud")
|
||
|
||
|
||
@app.post("/webhook")
|
||
async def handle_webhook(
|
||
request: Request,
|
||
x_nextcloud_talk_signature: Optional[str] = Header(None, alias="X-Nextcloud-Talk-Signature"),
|
||
x_nextcloud_talk_random: Optional[str] = Header(None, alias="X-Nextcloud-Talk-Random"),
|
||
):
|
||
body = await request.body()
|
||
|
||
if x_nextcloud_talk_signature and not verify_signature(body, x_nextcloud_talk_signature, x_nextcloud_talk_random):
|
||
log.warning("Invalid webhook signature")
|
||
raise HTTPException(status_code=401, detail="Invalid signature")
|
||
|
||
try:
|
||
data = json.loads(body)
|
||
except json.JSONDecodeError:
|
||
raise HTTPException(status_code=400, detail="Invalid JSON")
|
||
|
||
log.info(f"Received webhook: {json.dumps(data, indent=2)[:500]}")
|
||
|
||
actor = data.get("actor", {})
|
||
actor_type = actor.get("type", "")
|
||
actor_id_full = actor.get("id", "")
|
||
|
||
if "/" in actor_id_full:
|
||
actor_id = actor_id_full.split("/", 1)[1]
|
||
else:
|
||
actor_id = actor_id_full
|
||
|
||
obj = data.get("object", {})
|
||
message_id = obj.get("id")
|
||
content_str = obj.get("content", "{}")
|
||
try:
|
||
content = json.loads(content_str)
|
||
message_text = content.get("message", "")
|
||
except json.JSONDecodeError:
|
||
message_text = content_str
|
||
|
||
target = data.get("target", {})
|
||
conversation_token = target.get("id", "")
|
||
|
||
if actor_type not in ("users", "Person"):
|
||
log.info(f"Ignoring non-user actor: {actor_type}")
|
||
return JSONResponse({"status": "ignored", "reason": "not a user message"})
|
||
|
||
is_direct_message = False
|
||
|
||
bot_mentioned = False
|
||
clean_message = message_text
|
||
|
||
escaped = re.escape(BOT_NAME)
|
||
mention_patterns = [
|
||
rf'@"?{escaped}"?\s*',
|
||
]
|
||
|
||
for pattern in mention_patterns:
|
||
if re.search(pattern, message_text, re.IGNORECASE):
|
||
bot_mentioned = True
|
||
clean_message = re.sub(pattern, '', message_text, flags=re.IGNORECASE).strip()
|
||
break
|
||
|
||
if not is_direct_message and not bot_mentioned:
|
||
log.info("Ignoring message in group chat without mention")
|
||
return JSONResponse({"status": "ignored", "reason": "not mentioned in group chat"})
|
||
|
||
if bot_mentioned:
|
||
message_text = clean_message
|
||
|
||
if ALLOWED_USERS and actor_id not in ALLOWED_USERS:
|
||
log.warning(f"User {actor_id} not in allowed list")
|
||
await send_reply(
|
||
conversation_token,
|
||
"🚫 Du bist nicht berechtigt, diesen Bot zu nutzen.",
|
||
reply_to=message_id
|
||
)
|
||
return JSONResponse({"status": "rejected", "reason": "user not allowed"})
|
||
|
||
if not message_text.strip():
|
||
return JSONResponse({"status": "ignored", "reason": "empty message"})
|
||
|
||
log.info(f"Processing message from {actor_id}: {message_text[:100]}")
|
||
|
||
if message_text.strip().lower() in ("hilfe", "help", "?"):
|
||
help_text = f"""🤖 **{BOT_NAME} Bot Hilfe**
|
||
|
||
Schreib mir einfach eine Nachricht und ich antworte dir.
|
||
|
||
**Nutzung:**
|
||
• In Gruppenchats: @{BOT_NAME} gefolgt von deiner Frage
|
||
|
||
**Befehle:**
|
||
• `hilfe` oder `?` – Diese Hilfe anzeigen
|
||
|
||
Modell: `{MODEL_NAME}` @ `{MODEL_BASE_URL}`
|
||
Der Bot merkt sich die letzten Nachrichten pro Raum (bis zum Neustart)."""
|
||
await send_reply(conversation_token, help_text, reply_to=message_id)
|
||
return JSONResponse({"status": "ok", "action": "help"})
|
||
|
||
messages = build_messages(conversation_token, message_text, actor_id)
|
||
response = await call_model(messages)
|
||
|
||
history = conversations.setdefault(conversation_token, [])
|
||
history.append({"role": "user", "content": f"[{actor_id}] {message_text}"})
|
||
history.append({"role": "assistant", "content": response})
|
||
|
||
if len(history) > MAX_HISTORY * 2:
|
||
del history[: len(history) - MAX_HISTORY * 2]
|
||
|
||
await send_reply(conversation_token, response, reply_to=message_id)
|
||
|
||
return JSONResponse({"status": "ok"})
|
||
|
||
|
||
@app.get("/health")
|
||
async def health():
|
||
return {
|
||
"status": "ok",
|
||
"nextcloud_url": NEXTCLOUD_URL,
|
||
"model_base_url": MODEL_BASE_URL,
|
||
"model_name": MODEL_NAME,
|
||
"bot_name": BOT_NAME,
|
||
"allowed_users": ALLOWED_USERS if ALLOWED_USERS else "all",
|
||
"max_history": MAX_HISTORY,
|
||
}
|
||
|
||
|
||
if __name__ == "__main__":
|
||
import uvicorn
|
||
uvicorn.run(app, host="127.0.0.1", port=8086)
|